tfrs.examples.movielens.evaluate

Evaluates a Movielens model on the supplied datasets.

tfrs.examples.movielens.evaluate(
    user_model: tf.keras.Model,
    movie_model: tf.keras.Model,
    test: tf.data.Dataset,
    movies: tf.data.Dataset,
    train: Optional[tf.data.Dataset] = None,
    k: int = 10
) -> Dict[Text, float]

user_model User representation model.
movie_model Movie representation model.
test Test dataset.
movies Dataset of movies.
train Training dataset. If supplied, recommendations for training watches will be removed.
k The cutoff value at which to compute precision and recall.

Dictionary of metrics.